Many conventional blood cell isolation procedures include a preliminary bulk separation of erythrocytic and granulocytic components by density-gradient sedimentation. Density-gradient separation relies on small differences in the density of different cell types causing them to segregate at different levels in a fluid medium of variable density. Differences in density between the cell types can be small, and individual cells types can be heterogeneous in size and density. Consequently, particular cell types can become distributed throughout a density-gradient medium rather than precisely segregating at a discrete area in the density medium. This phenomenon can result in poor recovery of desired cells and/or contamination with undesired cell types. In procedures that enrich for rare blood cell types such as hematopoietic progenitor cells, density-gradient sedimentation generally results in poor yields. For example, using conventional density-gradient methods to isolate progenitor cells (e.g., CD34+ hematopoietic stem cells) from umbilical cord blood reportedly results in a significant loss of the desired stem cells. See e.g., Wagner, J. E., Am J Ped Hematol Oncol 15:169 (1993). As another example, using conventional density-gradient methods to isolate lymphocytes reportedly results in selective loss of particular lymphocyte subsets. See e.g., Collins, D. P., J Immunol Methods 243:125 (2000).
Increasing the recovery of rare cell types from donor tissue could dramatically improve the success of transplant and immune therapies (e.g., bone marrow transplants, stem cell-based gene therapy, and immune cell therapy), the success of which apparently is related to the actual number of the cells being used for therapy